The Database of Interacting Proteins (DIP) is the original and one of the most extensive databases of information on interacting proteins. It was founded on the assumption that macromolecular interactions are at the basis of all biomedical science, and that researchers in many branches of biology and medicine can benefit from access to an accurate repository, documenting what is known of protein interactions. DIP has forged the International Molecular Interaction Exchange Consortium (IMEx) with the two principal European interaction databases to cooperate in curation of literature information on interactions and with the SwissProt bioinformatics group. IMEx has established formats and standards for data curation, now endorsed by HUPO (the Human Proteome Organization). To continue its expansion, DIP with its IMEx partners and journal editors, is developing a procedure for direct deposition of experimental data by primary authors. Accuracy of interaction data being essential, the DIP methods for the assessment of data quality will be extended to define a Core set of highly reliable interaction data for each organism. The Core interactions will permit construction of networks of interacting proteins, useful to cell and systems biologists. From its inception, DIP has been guided by a program of research into methods for extracting knowledge of networks from genomic and proteomic data and methods for annotation of protein function. This research has resulted in the ProLinks database of functional linkages between proteins, the ProKnow server for annotating "unknown" protein sequences and structures, the method of Generalized Functional Linkages (GFL) for supplying functional annotations to proteins based on their networks of interactions. New depositions to DIP will be used to evaluate the effectiveness of inferred annotations from ProKnow and ProLinks. Thus, the challenge of curating DIP stimulates discovery of new methods which in turn enhance the quality and usefulness of DIP data, which in turn provide validation for the new methods. In the new grant period, the inferred linkages and annotations will be combined with DIP to offer researchers "one stop information" about the interactions and functional context of their proteins of interest. Every interaction, both experimental and inferred linkages, will be accompanied by its uncertainty value. A new Web interface is currently being implemented to ease user access to all this information. This groundwork should allow DIP to supply information on hundreds of thousands of interactions in the coming years. PUBLIC HEALTH RELEVANCE: The Database of Interacting Proteins (DIP) documents the discoveries of tens of thousands of scientists, described in thousands of research articles, about the interactions of proteins that form the basis of all biomedical science. DIP is a free and open community resource for research physicians and scientists. DIP takes special care in documenting the accuracy of the data it contains.